2018 ASEE Annual Conference & Exposition

This project is funded through the NSF Improving Undergraduate STEM Education initiative and seeks to enhance undergraduate computer science (CS) education by teaching computational creativity in both CS and non-CS courses. Computational creativity integrates computational thinking and creative thinking so that each can be used to enhance the other in improving student learning and performance in class. Whereas computational thinking brings a structured and analytic approach to problem-solving situations, creative thinking introduces novelty and innovative, non-standard solutions.
Computational creativity has been incorporated into numerous undergraduate CS and non-CS courses through a series of Computational Creativity Exercises (CCEs). Previous research indicates CCEs enhance students learning of computational thinking principles and boost course grades, even after controlling for students’ general academic achievement and individual differences in motivation and self-regulation. The effect of CCEs on achievement has been observed for both introductory and upper-level courses and for both men and women.
The present quasi-experimental study examined the impact of CCEs on students’ (N = 670) retention in CS courses. CCEs were incorporated into 100- and 200-level CS courses at a single university during the fall of 2015. Students in these classes consented to having their course enrollment data collected for the following three semesters (retention semesters). Students in 100- and 200-level CS courses during the fall of 2014 and spring of 2015 semester also consented to having their ongoing course enrollment data collected and were used as a comparison group. The impact of the CCEs on retention was tested through chi square analysis for each of the three retention semesters. Results indicate students in the implementation semester courses were more likely to continue taking CS in each of the three retention semesters (Cramer’s Vs from 0.144 to 0.193), and the effect was even greater when comparing students who did no CCEs (in either the implementation or control groups) and those who completed at least two CCEs (Cramer’s Vs from 0.153 to 0.254). Thus, these results suggest that in addition to the impact CCEs have on achievement, CCEs also increase the likelihood that CS students will continue to take CS courses.